Ensaios sobre eleições, financiamento de campanha, ideologia e grupos de interesse.

Abstract

This Thesis is composed of three essays that seek to infer and explain the relationships between elections, politicians, and their respective campaign funders. In addition, the products arising from this interaction are explored. In modern democracies, political representatives are chosen to meet the interests of the population, however, the political platform adopted by them is often influenced and biased in favor of the groups that supported their electoral campaigns. The first essay aims to identify the relationship between such groups of donors and the expenditure on public goods and services of the municipalities. What is the relationship between the funding profile of politicians in a municipality and local public spending? The donors were separated into two groups, Legal Entities (PJ) and Individuals (PF). The approach of the median voter model was used, which incorporates the variables that denote the different sources of demand variation for local public expenditures, plus the donation percentages of the PF and PJ groups of all candidates elected in 2008. The hypothesis of a linear relationship between these variables and a model with a threshold effect is estimated through the methodology proposed by Hansen (2000). The variable used to determine this threshold effect is GDP per capita. As the main results, the positive relationship between expenditures, current or capital, and the percentage donated by PJs to elected politicians stand out. Except in the case of the poorest municipalities, where the relationship between capital expenditures and PJ is insignificant. In turn, the coefficient of the PF variable is significant and positive in municipalities with lower GDP per capita and negative in those with higher income levels. In interactions between government and voters, there is a strong asymmetry of information. The government does not know what the citizens want, and the citizens, in turn, do not know what the government or opposition has done, is doing, or should do to serve their interests. The cost of comparing the policies of competing parties to the government is high. Furthermore, citizens do not always have enough information to assess the differences between such policies. Therefore, many voters find party ideologies useful insofar as they lower this cost. If the voter discovers a correlation between the ideology of each party and its policies, he can vote rationally by comparing ideologies rather than policies. The second Essay, based on the theory of the spatial vote developed by Downs (1957), tries to answer the following question: What is the ideology of Brazilian politicians? The main objective is to create a variable that can serve as a representation of ideology for politicians and, in a complementary way, for their respective parties. In addition, determine how many dimensions are important to explain the positioning of these parliamentarians. Once the ideological variable was defined, the main themes or areas that define it were inferred. Finally, an exercise in party reclassification based on the estimated ideology was proposed. The object of study was composed by the Federal Deputies of the 55th Brazilian legislature that took place from 2015 to 2018. In this research, the W-NOMINATE (Poole and Rosenthal, 1985), Optimal Classification (Poole et al., 1997) methods were used, as well as Principal Component Analysis. As the main results, it is highlighted that two spatial dimensions seem to explain most of the votes made. The first dimension, with greater explanatory power, is seen as ideological, while the second consists of the Government x Opposition relationship. It was identified that the ideology is linked to Public Finance and Budget issues, opposing, mainly, the groups formed by PT, PC do B, and PSOL to the group formed by PSDB, DEM, PP, SD, CIDAD, MDB, PTB, and PSC. The proposed reclassification exercise shows that taking into account the estimated ideal points, there should be between 2 and 7 political parties in Brazil. In August 2015, the Federal Chamber of Deputies approved the Proposed Amendment to the Constitution (PEC) 182/2007, known as the Mini-Political Reform PEC. Among the main points of the reform, the ban on corporate donations directly to candidates stands out. The last Essay tries to explain what led the Deputies to stop this source of funding. Based on the result of the vote on an Agglutinative Amendment (EMA) No. 22, which proposed the maintenance of direct donations made by companies to candidates, an econometric model was developed with the personal, political, and campaign financing characteristics of each parliamentarian who voted for or against the proposal. The model was estimated according to the Logit and Elastic Net method. The main hypothesis was that established politicians and those who did not depend on this source of funding would vote against the EMA, creating a barrier to the entry of new politicians, and ensuring less competition in subsequent elections, by reducing an important source of electoral funding. However, this theory has not been proven. The main determinants of votes were the Deputy's ideology, the more the right, the greater the chance of voting Yes, and the party's orientation, in case it was against the proposal, the probability of voting No increased.